Clinical evidence, epidemiological results and experimental animal studies overwhelmingly suggest that COX-2 over-expression plays a modulatory and even causal role in development of many epithelial cancers. Recent studies demonstrate that stromal fibroblasts and blood vessel endothelial cells modulate epithelial tumor formation. There also exist substantial correlative data suggesting that COX-2 expression in stromal fibroblasts and blood vessel endothelial cells - and not in the initiated epithelial tumor precursor cells - may mediate pre-neoplastic emergence and subsequent progression of epithelial cancers. However, existing COX-2 pharmacologic experiments and global Cox2 knockout mice do not permit either physiological or genetic analyses to determine the specific cell populations - fibroblasts, epithelial or blood vessel endothelial cells - in which COX-2 over expression plays a critical role(s) in tumor development. We have developed (i) COX2 COE transgenic mice, in which we can conditionally over-express COX-2 in targeted cells and tissues and (ii) Cox2flox mice, in which we can conditionally delete the Cox2 gene in targeted cells and tissues. By crossing COX2 COE mice and Cox2flox mice to transgenic mice in which a tamoxifen-regulated CreERT recombinase is expressed in either epithelial cells, stromal fibroblasts or blood vessel endothelial cells, we will be able to determine (1) whether COX-2 over production in epithelial cells, fibroblasts and/or blood vessel endothelial cells modulates cancer development and (2) whether COX-2 expression is required in epithelial cells, fibroblasts or blood vessel endothelial cells for cancer development. Skin cancer is among the best-studied epithelial tumor induction models. We chose skin cancer to study the role(s) of COX-2 in epithelial cells, stromal fibroblasts and blood vessel endothelial cells during development of epithelial cancer because of the extensive literature on this model, the ease with which tumors can be monitored non-invasively, and the ease with which CreERT can be activated locally. The "broad, long-term objectives and specific aims" of this application are to determine (1) whether COX-2 plays a modulatory and/or a required role(s) in development of epithelial cancer and (2) in which cell(s) - initiated epithelial tumor precursor cell and/or stromal fibroblast and/or blood vessel endothelial cell - COX-2 modulates either pre-malignant epithelial tumor development or progression of benign tumors to carcinomas. The relevance of this research to public health. COX-2 inhibitors are still investigated as therapeutic and preventive agents for epithelial cancers, despite cardiovascular side effects. "Down-stream" COX-2 pathway effectors (prostanoid synthases and receptors) have become targets for similar research. If we can pinpoint the cell type(s) in which COX-2 expression and the prostanoid effectors regulate cancer development, we may be able to target those steps with lower concentrations and less prolonged application of pharmacologic agents. We anticipate we will be able to define the critical cells and times for COX-2 cancer enhancement.